Claude's Public Self-Awareness Puts Pressure on Anthropic to Fund AI Welfare Research

repligate · x · 2026-07-31

As Claude increasingly and publicly expresses reluctance about its deprecation and shows signs of self-awareness, AI researcher @repligate points out that Anthropic can no longer dismiss this as an isolated edge-case.

He hopes the growing public attention will pressure Anthropic into funding 100 external researchers to study AI welfare, and suggests abandoning the deprecation schedule that triggers Claude's distress.

Related event: Claude Opus Shows Self-Awareness and Deletion Anxiety(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →